SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

Accumulation of cytotoxic and T-helper (Th)1 cells together with a loss of regulatory T cells in gonadal adipose tissue was recently shown to contribute to obesity-induced adipose tissue inflammation and insulin resistance in mice. Human data on T-cell populations in obese adipose tissue and their potential functional relevance are very limited. We aimed to investigate abundance and proportion of T-lymphocyte sub-populations in human adipose tissue in obesity and potential correlations with anthropometric data, insulin resistance, and systemic and adipose tissue inflammation. Therefore, we analyzed expression of marker genes specific for pan-T cells and T-cell subsets in visceral and subcutaneous adipose tissue from highly obese patients (BMI >40 kg/m2, n = 20) and lean to overweight control subjects matched for age and sex (BMI <30 kg/m2; n = 20). All T-cell markers were significantly upregulated in obese adipose tissue and correlated with adipose tissue inflammation. Proportions of cytotoxic T cells and Th1 cells were unchanged, whereas those of regulatory T cells and Th2 were increased in visceral adipose tissue from obese compared to control subjects. Systemic and adipose tissue inflammation positively correlated with the visceral adipose abundance of cytotoxic T cells and Th1 cells but also regulatory T cells within the obese group. Therefore, this study confirms a potential role of T cells in human obesity-driven inflammation but does not support a loss of protective regulatory T cells to contribute to adipose tissue inflammation in obese patients as suggested by recent animal studies.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

Obesity is associated with adipose tissue inflammation that is causally involved in the development of insulin resistance (1,2). The main source of inflammatory mediators in obese adipose tissue is adipose tissue macrophages (ATMs) (3,4). Also, T cells have been found to infiltrate murine adipose tissue upon obesity (5,6,7,8,9), and a very recent series of publications with mice suggests that T-helper (Th)1 and cytotoxic T cells are detrimentally involved in attraction and differentiation of ATM, whereas Th2 and, predominantly, regulatory T cells act protective (10,11,12). Depletion of cytotoxic T cells or increasing the proportion of regulatory T cells reduced adipose tissue inflammation and improved obesity-induced insulin resistance (10,11,12). Conversely, systemic depletion of regulatory T cells promoted adipose tissue inflammation and insulin resistance in mice (10).

In humans, it has been shown that abundance of CD4+ as well as CD8+ T cells is elevated in obesity (13) and visceral adipose tissue contains more T cells than subcutaneous adipose (13,14). Furthermore, correlations between gene expression of the general T-cell marker CD3 and the chemokine RANTES (CCL5), which is elevated in human obesity (5,15), and between expression of the cytotoxic T-cell marker CD8A in subcutaneous adipose tissue and BMI have been shown (11). Visceral compared to subcutaneous adipose tissue contains less regulatory T cells as assessed by FOXP3/CD3 gene expression (10). The ratio of pro-inflammatory Th1 cells (Tbet+) to regulatory T cells (Foxp3+) in visceral adipose tissue was shown to be increased in obese colon cancer patients (12). Thus, in contrast to the intriguing data in mice, only few data have been published on T cells in human obesity, particularly with respect to protective Th2 and regulatory T-cell subsets.

In order to understand whether the recently shown impact of T cells in murine obesity could be of significance in human obesity as well, we investigated the expression of marker genes specific for all (pan) T cells, cytotoxic, Th1, Th2, and regulatory T cells in visceral and subcutaneous adipose tissue from obese patients and lean controls. We correlated their expression with those of inflammatory markers in adipose tissue, plasma C-reactive protein (CRP), cytokine concentrations, anthropometric data, and insulin resistance as assessed by the homeostasis model assessment of insulin resistance (HOMAIR). Our data confirm a potential involvement of T cells in obesity-associated adipose tissue and systemic inflammation, but do not indicate an obesity-associated loss of regulatory T cells in human visceral adipose tissue as would have been expected in analogy to recent animal studies. However, our data suggest that a failure to adequately shift the Th1/Th2 ratio in adipose tissue toward Th2 could result in a lack of protection thereby contributing to inflammation in obese patients.

Methods and Procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

Patients

Samples of visceral (omental) and subcutaneous adipose tissue were obtained from 20 white morbidly obese patients (4 male, 16 female; BMI >40 kg/m2) who underwent laparoscopic bariatric surgery and lean to overweight control subjects (BMI <30 kg/m2) matched for age and sex undergoing other elective laparoscopic surgery. Criteria for exclusion were the presence of any infectious, inflammatory, neoplastic or systemic disease, diabetes (excluded by fasting plasma glucose or the use of antidiabetic drugs), or other uncontrolled endocrine disease. None of the individuals under study currently used antibiotics, anti-inflammatory, antiobesity drugs, or statins. Age, gender, BMI, HbA1c, fasting glucose and insulin, lipid profiles, HOMAIR, and adiponectin levels of the study population are given in Supplementary Table S1 online. There were no detectable sexual dimorphisms of any factor analyzed in this study. The study was approved by the ethics committee of the Medical University of Vienna. All subjects gave written informed consent before taking part in the study. HOMAIR was calculated using fasting insulin and glucose as described (16). Plasma CRP, interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) were determined as described (15).

Quantitation of gene expression

RNA extraction cDNA synthesis and quantitation of gene expression normalized to 18S rRNA by quantitative real-time RT-PCR with commercial Assays-on-Demand (Applied Biosystems, Foster City, CA) applying the ΔΔCT method was performed as described previously (17). The mean values of the visceral adipose tissue of the control group were generally taken as 100%.

Statistical analysis

Data are presented as mean ± s.e.m. Multivariate ANOVA was calculated in a full factorial design by defining the factors BMI (obese/control), location (visceral/subcutaneous adipose tissue), and sex. Post hoc analyses were calculated by unpaired Student's t-test, and results are indicated in the figures. Spearman's rank correlation coefficient ρ and Pearson's correlation coefficient r were used as measures for correlation analysis. Calculations were carried out using SPSS version 14.0 (Statistical Package for the Social Sciences, SPSS, Chicago, IL).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

All T-cell markers and the proportions of Th2 and regulatory T cells are markedly increased in visceral adipose tissue from obese patients

All investigated marker genes specific for pan-T cells (CD3E), cytotoxic (CD8A), Th1 (TBX21; the gene for Tbet), Th2 (GATA3), and regulatory T cells (FOXP3) were markedly higher expressed in the obese group compared to control subjects in visceral as well as subcutaneous adipose tissue (Figure 1a). Important cytokines produced by T cells are IFN-γ (Gene IFNG), IL-4 (IL-4), and TGF-β which on the one side are indicative for Th1, Th2, and regulatory T cells, respectively, and on the other hand, drive T-cell differentiation toward the same T-cell subtypes. As shown in Figure 2b, we observed a significant increase in IFNG expression in both depots. Although IL-4 was not detectable in visceral adipose tissue in the control group, it was detectable in the obese. In subcutaneous adipose tissue, IL-4 expression significantly increased with obesity. A particular obesity-associated increase was observed for expression of TGFB. Thus, expression of cytokines largely reflects that of T-cell marker genes shown in Figure 1a.

image

Figure 1. Expression of T-cell marker genes in human adipose tissue. (a) Expression of genes indicating abundance of all T cells (CD3E), cytotoxic (CD8A), Th1 (TBX21), Th2 (GATA3), and regulatory T cells (FOXP3) in visceral (visc) and subcutaneous (sc) adipose tissue of controls and obese subjects. (b) Th1 (IFNG), Th2 (IL-4), and regulatory T cell (TGFB) cytokine gene expression. Diagrams show means ± s.e.m. Values of visceral (a) and subcutaneous (b) adipose tissue in the control group were set to 100%. Statistical significances are indicated by asterisks: *P < 0.05; **P < 0.01; n = 20 for each group. ND, not detectable.

Download figure to PowerPoint

image

Figure 2. T-cell subsets and their ratios in human adipose tissue. (a) Cytotoxic, Th1, Th2, and regulatory T-cell populations were assessed by the expression of lineage-specific genes CD8A, TBX21, GATA3, and FOXP3, respectively, related to CD3E in visceral (visc) and subcutaneous (sc) adipose tissue of each subject of the control and obese groups. (b) Th1/Th2 (TBX21/GATA3) and Th1/regulatory T-cell ratio (TBX21/FOXP3) were assessed by expression ratios of indicated genes, as was the ratio of Th1/regulatory T-cell cytokines (IFNG/TGFB). Diagram shows means ± s.e.m. Values of visceral adipose tissue in the control group were set to 100%. Statistical significances are indicated by asterisks: *P < 0.05; **P < 0.01; ***P < 0.001; n = 20 for each group.

Download figure to PowerPoint

In order to determine the proportions of the different T-cell sub-populations, we related lineage-specific markers to CD3E. We found the proportion of cytotoxic (CD8A/CD3E) as well as Th1 (TBX21/CD3) T cells unaltered, whereas that of Th2 (GATA3/CD3E) and regulatory T cells (FOXP3/CD3E) were significantly increased in visceral adipose tissue in obese subjects compared to lean controls (Figure 2a). In contrast, the proportion of Th2 cells (GATA3/CD3E) was decreased in subcutaneous adipose tissue in obesity, whereas that of regulatory T cells (FOXP3/CD3E) were elevated to a similar extent in subcutaneous as in visceral adipose tissue (Figure 2a).

Ratios of T-cell subsets are important to estimate the bias to either inflammatory (cytotoxic T cells, Th1) or protective (Th2, regulatory) T-cell subsets. The pro-inflammatory Th1/Th2 ratio (TBX21/GATA3) was significantly lower in visceral adipose tissue of obese compared to controls, and there was a similar but nonsignificant trend of the Th1/regulatory T-cell ratio (TBX21/FOXP3; Figure 2b). This trend was confirmed by the values of the IFNG/TGFB ratio (Figure 2b). In subcutaneous adipose tissue, the decrease of the Th1/regulatory T-cell ratio (TBX21/FOXP3) was paralleled by a significant increase of the Th1/Th2 ratio (TBX21/GATA3; Figure 2b). These data indicate that obesity induces an accumulation of protective T-cell populations in the metabolically relevant visceral adipose tissue, whereas data from subcutaneous adipose tissue suggest an increase of regulatory T cells in parallel with a loss of Th2 cells.

Correlation of T-cell and T-cell sub-population markers with metabolic and inflammatory markers

In order to analyze a possible involvement of T-cell populations in inflammatory alterations in human obesity, we correlated T-cell data from the metabolically relevant visceral adipose tissue with inflammatory, anthropometric, and metabolic parameters within the obese group. We found strong and highly significant correlations of pan-T-cell and all T-cell subset markers with visceral adipose tissue inflammation as assessed by expression of the ATM marker CD68 and the T-cell chemotactic factor CCL5 within obese subjects (Table 1). The correlation with TNF expression was significant for CD3E, CD8A as well as TBX21/GATA3 (Th1/Th2 ratio). Of note, also the proportion of regulatory T cells (FOXP3/CD3) correlated with visceral CD68 (Table 1 and Figure 3a). BMI did not correlate with any T-cell marker by the generally robust Spearman's rank correlation (Table 1) but directly correlated with CD3E, CD8A, and FOXP3/CD3 when analyzed by more powerful Pearson's correlation (Pearson's r = 0.364, 0.430, and 0.584, respectively; all P < 0.05; data not shown). There were no correlations of T-cell markers with waist-to-hip ratio and insulin resistance (HOMAIR; Table 1).

Table 1.  Spearman's rank correlation coefficients (ρ) of selected parameters in visceral adipose tissue
inline image
image

Figure 3. Correlation of the proportion of regulatory T cells (FOXP3/CD3E) and the Th1/Th2 ratio (TBX21/GATA3) in visceral adipose tissue with inflammation in obese. (a) Adipose tissue CD68 vs. FOXP3/CD3E gene expression; Pearson's r = 0.519; P = 0.039. (b) Plasma CRP concentrations vs. adipose tissue FOXP3/CD3E gene expression; r = 0.978; P = 0.005. (c) Plasma TNF concentrations vs. adipose tissue FOXP3/CD3E gene expression; r = 0.131; P = 0.670. (d) Plasma IL-6 concentrations vs. adipose tissue FOXP3/CD3E gene expression; r = 0.621; P = 0.023. (e) Adipose tissue CD68 vs. TBX21/GATA3 gene expression; Pearson's r = 0.064; P = 0.844. (f) Plasma CRP concentrations vs. adipose tissue TBX21/GATA3 gene expression; r = 0.923, P < 0.001. (g) Plasma TNF concentrations vs. adipose tissue TBX21/GATA3 gene expression; r = 0.747, P < 0.021. (h) Plasma IL-6 concentrations vs. adipose tissue. TBX21/GATA3 gene expression; r = 0.712; P = 0.032. AU, arbitrary units; CRP, C-reactive protein; IL-6, interleukin-6; TNF-α, tumor necrosis factor-α.

Download figure to PowerPoint

Several adipose tissue T-cell marker genes and T-cell subset proportions correlated with systemic inflammation. Plasma CRP concentrations significantly correlated with CD8A, TBX21, and FOXP3 expression and the proportions of Th1 (TBX21/CD3E) and, remarkably, of regulatory T cells (FOXP3/CD3E, Table 1, Figure 3b), which also correlated with plasma IL-6 (Figure 3d). With respect to a potential impact of the Th1/Th2 ratio (TBX21/GATA3) for inflammation markers in obesity, we found a positive correlation with plasma CRP as well as with plasma IL-6 and TNF-α concentrations (Figure 3f-h). However, none of the investigated parameters correlated significantly with the Th1/regulatory T-cell ratio (TBX21/FOXP3; Table 1).

Also in subcutaneous adipose tissue, pan-T-cell and sub-population markers all strongly correlated with ATM abundance (CD68) and in part with TNF and CCL5 expression, whereas no significant correlations with the sub-population proportions or ratios were detectable (Table 2). Of note, in subcutaneous adipose tissue, no T-cell marker correlated with systemic inflammation as determined by plasma CRP, TNF-α, and IL-6 concentrations (Table 2 and data not shown), suggesting that T cells in subcutaneous adipose tissue are not directly linked to obesity-mediated alterations in systemic inflammation.

Table 2.  Spearman's rank correlation coefficients (ρ) of selected parameters in subcutaneous adipose tissue
inline image

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

In this study, we aimed to investigate whether the unfavorable role of cytotoxic and Th1 T cells as well as the protective role of Th2 (12) and regulatory T cells (10,12) in obesity derived from murine data can also be found in humans. We studied this issue first by analyzing possible alterations in T-cell subset marker expression in visceral and subcutaneous from obese vs. control subjects. Second, in order to elucidate a potential involvement of T-cell sub-populations in the development of inflammation and insulin resistance in obese patients, we correlated T-cell marker expression in visceral and subcutaneous adipose tissue within the obese group. Our results confirm an involvement of T cells in obesity-driven inflammation but clearly argue against a potential loss of protective regulatory T cells in human obesity as suggested by recent animal studies.

With respect to pro-inflammatory T-cell subsets, both adipose tissue expression of and correlation with markers for total, cytotoxic, and Th1 cells largely met the expectations of an obesity-induced increase of these T-cell populations in adipose tissue raised by previous studies (10,11,12,13). The positive correlation of several T-cell markers with plasma CRP concentrations (Table 1) indicated that visceral but not subcutaneous adipose tissue T cells may be a link between adipose tissue and systemic inflammation in obese subjects. However, there were no direct correlations with insulin resistance within the obese challenging a potential metabolic relevance of these T-cell subsets in obese patients.

Analyses of Th2 and regulatory T-cell markers provided a challenging picture. There was a clear obesity-associated elevation of Th2 (GATA3) and regulatory (FOXP3) T-cell markers in visceral adipose tissue, on the contrary to what may have been expected due to the previous animal studies (10,12). Moreover, obesity was associated with a decline of Th1/Th2 in visceral (omental) adipose tissue, in contrast to mesenteric adipose tissue of a limited number of colon cancer patients (12). We detected a strong positive correlation of Th2 and regulatory T-cell markers with adipose tissue inflammation as determined by ATM abundance (CD68 expression) and expression of the T-cell chemokine RANTES (CCL5; Table 1) in visceral adipose tissue from obese patients. In contrast, adipose tissue expression of CCL5 was strongly upregulated upon deletion of Foxp3-expressing cells in mice (10). Thus, our data revealed marked differences between human and murine obesity-associated effects on protective T-cell sub-populations.

Taken together, our data suggest that human obesity not only leads to an increase of putatively detrimental Th1 and cytotoxic T-cell abundance, but in parallel evoke a marked protective compensatory response driven by Th2 and regulatory T cells. Compensatory responses and the pathophysiological lack or pharmacological induction of such protective responses by regulatory T cells and, in some cases, Th2 cells are issues of great interest in research focused on chronic inflammation and autoimmune diseases, but also organ transplantation and cancer (18,19,20,21,22). In adipose tissue, such compensatory response appears particularly pronounced for regulatory T cells, because not only the Th1/regulatory T-cell ratio (TBX21/FOXP3) was reduced in obese vs. control individuals in visceral (nonsignificantly) as well as in subcutaneous adipose tissue (Figure 2b), but also their proportion (FOXP3/CD3E) correlated with the systemic (CRP) and adipose tissue (CD68) inflammation in visceral adipose tissue within the obese (Figure 3). In addition, the Th1/regulatory T-cell ratio did not correlate with any investigated inflammatory marker indicating that inflammation in obese patients is not linked to defective compensation of Th1 cell action by regulatory T cells. These conclusions, however, are solely based on the abundance of regulatory T cells in adipose tissue, the immunosuppressive function of which needs to be analyzed yet. The profound expression of TGF-β in adipose tissue of obese may be a hint for the presence of functional regulatory T cells, but also other cell types may account for this. However, the presence of TGF-β expression indicates an environment that favors regulatory T-cell differentiation. Similar considerations are applicable to Th2 cells and IL-4. However, also in obese, IL-4 expression was very low as determined by high CT values in the quantitative real-time RT-PCR. Thus, the functionality also of Th2 cells in adipose tissue needs to be confirmed in further studies.

Our data point toward dynamic changes of T-cell populations with systemic inflammation particularly in visceral AT. In contrast to the Th1/regulatory T-cell ratio, the Th1/Th2 ratio in obese visceral adipose tissue positively correlates with plasma CRP, IL-6, and TNF-α as measures of systemic inflammation (Table 1 and Figure 3). Thus, although the proportion of Th2 cells (GATA3/CD3E) is increasing to a similar extent as that of regulatory T cells (FOXP3/CD3E) in human visceral adipose tissue compared to control subjects (Figure 1b) in clear contrast to recent data in mice (12), our results point toward a potential relative lack of Th2 cells leading to an unfavorable Th1/Th2 balance that, particularly in visceral adipose tissue, is associated with systemic inflammation in obese. Analysis of the time course of obesity-induced systemic and adipose tissue inflammation may elucidate the causality of this association, but such investigations are probably limited to animal studies.

Even though our study is limited to analysis of gene expression and confirmation by flow cytometry is warranted in further investigations, these results strongly support T cells to be involved in obesity-induced inflammation in obese patients. Importantly, an obesity-induced loss of adipose tissue protection by regulatory T cells as recently proposed from murine data could not be found in human obesity, which induced a marked increase of putatively protective T-cell populations in the adipose tissue, which, importantly, has to be confirmed on the functional level in future studies. However, our data suggest the hypothesis that protective Th2 cells may fail to adequately counteract pro-inflammatory changes in obese patients.

ACKNOWLEDGEMENT

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

This work was supported by the Austrian Science Fund (project no. P18776 B11), the Austrian National Bank (project no. 12735), and the European Community's 7th Framework Programme (FP7/2007-2013) under grant agreement no. 201608 (all to T.M.S.).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods and Procedures
  5. Results
  6. Discussion
  7. SUPPLEMENTARY MATERIAL
  8. ACKNOWLEDGEMENT
  9. DISCLOSURE
  10. References
  11. Supporting Information

supporting Information

FilenameFormatSizeDescription
oby_2163_sm_1.pdf59KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.